ss won't run with warm engine

[93Leg-c]

I thought I had started a thread on this some time ago but I can't find it.

Background: '93 ss with AT that has been basically sitting for about five years. Rats got into the engine bay (lots of droppings) with one spark plug wire slightly chewed. The ss had driveability issues this past September -- wouldn't idle (engine would stumble, stall, and die) and would sputter and die when attempting to drive it. Pulled codes and found VSS, O2 sensor, inhibitor switch, and parking switch. Replaced O2 sensor and VSS and the car seemed to have solved idling issues but not driveability issues. Spark plug wires replaced with used ones that were working previously (but now that I think about it, I did not check them out to make sure they were in good working order). I had to leave off working on the car until this past week when I drained 15 gallons of old gas, replaced spark plugs, fuel filter, all fluids, fuel pump, and made a new knock sensor wire because the original one broke due to being brittle (but it did not spec out at 0 ohms but .2 ohms).

Current issue: Car starts instantly, idles perfectly, engine sounds healthy and strong, revs easily, but only when engine is cold. Drives fine when engine is cold. Once it warms up, it is difficult to drive -- it sputters and stalls. Whether I feather the gas or give it a lot, it either kills the engine or gets very close to it. I have to let it idle for a while then take the rpms up to 3K and leave it there for about 15 seconds, bring the rpms back down to about 1500 then put it in Drive and drive a few house lots before it starts stalling again. The better way is to turn the engine off, start it up and immediately bring it up to 3K rpms, bring the idle back down to 1500, put it in Drive and then I can drive several blocks or more. For some reason, turning the engine off seems to help it.

Current CEL codes pulled: 22 (knock sensosr), 31 (throttle sensor), 32 (O2 sensor-- but only when engine is cold), 33 (vehicle speed sensor -- but only when the engine is warm), 51 (inhibitor switch), and 52 (parking switch). The Power light for the TCU also flashes 16 times upon startup but I'm not able to pull the codes since I cannot drive the car after the engine warms up. Oh, the tach and speedometer needle also vibrate; it used to be steady and this is the first time they vibrate.

Which of those codes affect the driveability issue? I'm pretty sure the throttle sensor and VSS has an effect but what about the rest? A side question on this is: If the speedometer cable (to which the VSS is attached to) is not plugged into the speedometer all the way, will that cause the CEL to come on as well as have the speedometer and tach needles to vibrate? How do I know if the cable is inserted all the way in? Also, where is the parking switch located?

Any input is really appreciated! Thanks!

[Legacy777]

Len,

The drivability issues when warmed up sound like a coolant temp sensor. Honestly, I would start there before trying to tackle the other codes. it's relatively cheap, around $20 and is the biggest culprit when it comes to temperature dependent issues.

Does the speedometer work? If so, the VSS code is likely a false code....unless there are wiring issues between the gauge cluster and ECU.

[93Leg-c]

Thanks Josh! I will do that -- get a new coolant temperature sensor tomorrow.

Oh, by the way, why didn't that come up in the codes?

[93Leg-c]

To answer your question, yes the speedometer does work. However, the needle now vibrates or flutters (same with the tach needle). It never did that before. What could be causing that?

So if there are wiring issues between the gauge cluster and ecu, do I have to pull the dash to check the wiring?

[Legacy777]

A failing coolant temp sensor doesn't typically throw a code in these cars. This is because the ECU's are for a lack of a better term, dumb. They do not have the logic built into them like the new ECU's that the car should "warm up" within a certain amount of time.

I can't guarantee that the coolant temp sensor is the problem, but engine temperature operating issues is usually an issue with the coolant temp sensor.


Replace the coolant temp sensor before moving onto the next issue. You don't need to pull the dash to check the wiring, but you'd need to pull the black trim around the gauge cluster and then pull the gauge cluster. If both the speedo & tach are doing it, it may be a ground issue.

[93Leg-c]

I've installed a new coolant temp sensor but haven't started the ss yet because one of the nipples in the intake resonator box was broken. So, it's in the JB Kwik Weld 24-hour curing stage. I'll connect everything tomorrow morning and see what happens after I start it. (As a side, this is the coldest it's been in February. Lots of snow down to the base of the mountains. So, it's going to be chilly for me working on the ss tomorrow.) I'll report back with the results.

Regarding the grounding issue, is there any particular ground wire I should check or do I need to check them all? And it's good to know I don't have to pull the dash but just the gauge cluster, if needed.

Thanks for all your help, Josh. It's very much appreciated!

[Legacy777]

Hopefully the coolant temp sensor will fix your problem. Keeping my fingers crossed.

[93Leg-c]

Well, the nipple broke off when installing the resonator box. The line from the broken nipple leads to the bottom of the wastegate control solenoid. Does it have to be hooked up and working properly for me to take the ss for a test drive?

[Legacy777]

That is the vent line for the wastegate solenoid. You could probably run it without that going back into the intake, but before trying to troubleshoot any other iseues, I'd get that fixed so the engine is running on the proper air & fuel mixture.

How would you plug the hole on the resonator box?

[93Leg-c]

Thanks for the advice. I thought that would be the answer but I thought I'd ask anyway because if it was a possibility to run the engine, then I'd run the ss to see if the coolant temperature sensor got rid of any or all of the CEL codes. But since I won't be starting the engine tomorrow, I'll work on other things on the ss, like installing the rear GR-2 struts (well, actually, they are the Excel-G struts), working on the brakes, etc.

If I had to plug the resonator box hole temporarily (which would disrupt the A/F mixture), it would be with a screw. But . . .

I have a repair going on on the resonator box. I went to the auto parts store and got 1/4" plastic (or whatever it's made of) inline barbed fittings. The fittings have barbs on both ends with "finned" support in the middle. I cut off one end of the fitting but left about 3/16" of it to insert into the resonator box hole that I drilled out to snugly accommodate the fitting. The two "fins" rest on the resonator box and give the barbed fitting added stability. I applied plastic weld (it has a 3500 psi tensile strength -- more than JB Weld or epoxies) to the side of the resonator box hole and the fitting, inserted the fitting, then globbed a whole bunch of the plastic weld over the outside bottom of the fitting where the hose would not touch. Cure time is 24 hours which means tomorrow evening at about 7 PM my time. So, on Thursday I will assemble the parts and see if it works.

What I will also do to put less stress on the repair is to lengthen the hose from the resonator box to the wastegate solenoid by adding more tubing and connecting it with another inline barbed fitting. I'm doing that because the resonator box usually ends up getting twisted when removing and reinstalling it. The pressure or force of that twist is enough to possibly break the repair. I'm figuring with the extra long hose and connection in between, I can disconnect that line at the inline fitting when pulling off the resonator box to eliminate the twisting force on the repair and upon reinstalling the resonator box, that hose will be connected the last at the inline fitting.

Is my reasoning sound? Will the extra length of tubing disrupt anything in the system?

EDIT: The first repair with the JB Kwik Weld broke off for possibly several or all of the following: The repair nipple had a base to it and was just resting on top of the resonator box and did not extend into the box; I may not have put enough of the quik weld on the outside of the fitting; the fitting broke away from the box when I attempted to pull the hose out of the barbed fitting.

[Legacy777]

What you've got laid out sounds reasonable.

[93Leg-c]

Hopefully the coolant temp sensor will fix your problem.

Well, the coolant temp sensor did not remedy the issue(s).

The engine idles fine whether it's cold or warm but when driving the car, once I hit 2,000 rpms, the engine starts to cut out like it's fuel starved. The engine will then have issues idling and revving. The engine usually dies. If I prevented it from dying by feathering the gas, it won't run smoothly nor rev much higher than where it is at that moment. But if I turn off the ignition then restart the car, I can rev it fine. I can then drive it, but only below 2,000 rpms.

The codes that now come up are:
22 knock sensor
31 throttle sensor
33 vehicle speed sensor
35 canister purge solenoid valve
51 neutral / inhibitor switch
52 parking switch

On my final testing when parked I noticed when I was revving the engine that it made a metallic sound, like something heavier than heat shields were loose. It didn't sound like the pre-ignition sounds and knocking that I'm familiar with but I'm wondering if it is pre-ignition. The metallic pinging/ clear ringing sound usually started at around 3,000 to 3,250 rpms but once it came out at around 2,500 rpms.

As a review, the VSS is new in the ss. It doesn't have even one mile of use on it. The knock sensor wire is new; however, it spec'd out at .2 ohms instead of 0 ohms.

Is there a possibility that the cat(s) are plugged? How do I test for clogged cats? Is it also possible that some wiring is chewed where I haven't discovered it yet?

Any help and suggestions are appreciated!

[Legacy777]

Len,

Well that sucks the CTS didn't fix the problem.

Quick question for you, have you checked whether the canister purge solenoid is working? If it's not, you might try replacing that or putting a resistor in it's place so the ECU thinks it's working properly. The reason I mention this is because some cars, which may be CA identified cars might not clear other codes if the canister purge solenoid isn't working. I guess another option you could look at to see is if your car has a pin/wire at the CA/Fed pin on the ECU.

The chances of all those codes remaining are pretty slim. If you've verified that the wiring and sensor are in spec, the only thing remaining is the ECU.

Just a curiosity, but I wonder if the MAF sensor or MAF sensor wiring might be suspect.

[93Leg-c]

I was up into the wee hours of the morning reading other threads on the bbs. I think it was Alphius posting on a similar issue that the connectors on the harnesses next to the battery need to be checked. So, all three were disconnected and sprayed down with wiring spray cleaner (I forgot the exact name of it) and some dielectric grease was used on each of the pins. Also checked the TPS wiring and the wires just before going into the connector to the TPS looked a bit suspect (kinda worn looking). The continuity of each wire tested good but I wrapped each one individually with electrician's tape.

Then, the ss was started and taken for a spin. It was MUCH improved! I could get it up to 45 mph with no problems, and of course, over 2K rpms. However, the knock sensor code (22) started flashing as I was driving (yes, I was driving in D-check mode) and the engine started warming up. It started sputtering a little and especially going up hill. I had a friend follow me and he said black smoke was coming out of the exhaust when the car started sputtering and slowing down.

At least all but the knock sensor codes are now gone. I'm heading to Fry's Electronics to get some metal-shielded, stranded wire since the one I made up spec's out at .2 ohms rather than 0 ohm.s

I was going to ask about the MAF possibly malfunctioning. But since none of the codes for that came up I haven't done anything with it.

But, as usual, please free to offer suggestions. I appreciate any and all of them.

[93Leg-c]

Just a quick update: I never did go to Fry's because I stopped in at the local Goodwill and found some Monster cables for sale for $5.99. I Googled the composition of them and found that the primary wire is stranded and it has stranded wire and aluminum sheathing for shielding. It spec's out at 0 ohms so I bought it. The primary wire is a bit larger than the original, and not thinking about it, soldered it too closely to the pins so it doesn't fit into the connectors. I'll have to re-do the ecu ends tomorrow.

[Legacy777]

Len,

Glad the cleaning of those connectors helped sort out those other codes. I guess let us know if getting the new knock sensor wire in fixes the knock sensor code and the issues once warmed up.

[93Leg-c]

Well, the new sensor wire seems to have fixed the 22 code (knock sensor). The ss runs better. HOWEVER, it starts stuttering at about 2200 rpms when the engine gets warm. Once I start up an incline, the power drops off. If I feather the gas, I can keep the car driving but only about 10-20 mph. As long as I am in gear in D-check mode, code 51 comes up; once I put it in N or P, the CEL flashes regularly indicating there are no issues. Code 51 was one of the codes that came up but went away with the previous knock sensor wire. Code 51 is the neutral/ inhibitor switch.

What does the neutral / inhibitor switch do?

One of my friends suggested it could be a clogged catalytic converter. But it revs fine in N or P whether the engine is cold or warm.

Anyone has suggestions?

[93Leg-c]

I did more testing on the ss today. I found that even when the engine is cold when it reaches about 2200 rpms, the engine starts stalling out. So, no longer is the stumbling/ stalling temperature sensitive. And It also doesn't matter what gear I'm in, once it reaches 2200 rpms, the stalling starts.

I had two friends with me today trying to figure out the problem. We blocked open the wastegate then test-drove the ss. There was no longer any stalling at 2200 rpms but usually about 2800 rpms the symptoms started but also less severe. Only once did the stumbling start at around 2500 rpms.

We returned the wastegate function to normal and then test-drove the ss again. It went back to stalling at 2200 rpms and when I tried getting up to 2800 rpms, the bucking, stalling, stumbling was definitely more severe.

Is our conclusion correct that probably the cause of stumbling comes from some sort of backpressure due to malfunction of the turbo or clogged cats? Three guys I know suggested clogged cats as they have experienced it in their cars though the symptoms are not exactly the same.

I have access to a wrx downpipe. Can I just bolt that onto my turbo and take if for a short test drive to see if the bucking, stalling, stumbling goes away? Does anyone know what has to be done to make it a permanent fit if it does work? One of my friends is a welder and he may be able to help me out.

Any input is appreciated!

[Legacy777]

I think swapping in just the WRX downpipe is probably a good next step. As for permanent fitment, I think it just needs to be shortened.

[Legacy777]

Regarding the inhibitor switch, it's what tells the TCU and ECU what gear the AT is in.....as well as is connected to the linkage in the transmission that shifts gears.

[93Leg-c]

I had to leave the ss with a friend who's a mechanic since I had to get back to work and the ss is still not running right.

I just started a new thread since it's a specific question but it may be related: http://bbs.legacycentral.org/viewtopic.php?f=19&t=51584